We propose to conduct detailed analyses of both hormonal and dietary frisk factors for ovarian cancer using data collected since 1976 in the prospective Nurses' Health Study cohort. From 1976-2004, we expect 737 incident ovarian cancer cases to be confirmed in total. Specifically, we propose to evaluate body mass index, waist-to-hip ratio, antioxidant intake, time since last use of oral contraceptives and duration of postmenopausal hormone use in relation to ovarian cancer risk. In addition, using a nested case-control design, we will examine relationships between plasma insulin-like growth factor I (and its binding proteins) and ovarian cancer using blood samples collected from 32,826 cohort members in 1989-90. Using germline DNA from either archived buffy coat, buccal cell specimens or tissue (the latter two to be collected as part of this program project), we propose to evaluate ovarian cancer risk in relation to several specific genetic polymorphisms and gene- environment interactions (galactose-1-phosphate uridyl transferase [GALT] and lactose intake; GAG repeats in the androgen receptor gene and oral contraceptive use). In addition, we propose to establish a bank of tumor tissue from women with incident ovarian cancer, and to assess several ovarian cancer risk factors according to over-expression of mutant p53 protein in the tumor tissue. Overall, the prospective nature and large size of the cohort, the high follow-up rates over more than 24 years, the detailed and repeated measured exposure data, and the availability of both germline DNA and tumor tissue from a subset of the cohort provide a unique opportunity to test several important hypotheses related to risk of ovarian cancer.